Contingencies of localization in literary theory

Statements in literary theory can be understood as acts of disjunction, since they are based in one way or another on the isolation of what in a given literary text is actually considered suitable, appropriate or conducive to a theoretical utterance. As the introduction to this issue states, the very “accuracy of critical theories” can be “measured by the fact that they can tell the essential from the accidental, and separate the significant and regular features of a work or genre from what could simply be described as ‘noise’.” However, it is rare to find more specific methodological information in theoretical texts about how such an isolation or disjunction is carried out. Rather, there seems to be a high degree of intuition and, along with it, arbitrariness and randomness at play here, even in theories that are very reflective in the explication of their presuppositions and in the manner of their argumentation. Thus the question can be raised if and how literary theory, in the selection of ‘theorizable’ parts (or sections, or passages) of texts, is particularly “capable […] of managing the role played by chance’.”

In this paper I will focus on the aspect of localization. Where are the theoretically relevant parts and passages to be found in the literary texts? Where do these passages begin and end? How do they stand out from their less meaningful surroundings? I will address these problems, first, in the hermeneutical considerations of Friedrich Schleiermacher, who, on the one hand, has a holistic claim to textual understanding and, on the other, concedes that the whole can always be grasped only in its individual parts. From there, secondly, I turn to the concepts of ‘places of indeterminacy’ and ‘voids’ in reception aesthetics. Finally and thirdly, I discuss an attempt by Jacques Derrida to discuss a text in its completeness by representing the totality of its ‘places.’ In all three cases, the question is to what extent the fortuitousness and intractability of textual localization can be theoretically ordered and regulated.¹

In one of his lectures on hermeneutics presented to the Berlin Royal Academy of the Sciences in 1829, Friedrich Schleiermacher formulated the “hermeneutic principle” that “just as the whole is understood from the individual, so also the individual can only be understood from the whole” (Schleiermacher, 2002, p. 625).² With this fundamental statement, he refers to the dominant hermeneutics textbook of his time, the Grundlinien der Grammatik, Hermeneutik und Kritik by the classical philologist Friedrich Ast, published in 1808. While Schleiermacher accuses his colleague of “obfuscation and floating” (“Nebelei und Schwebelei”) because Ast claims that the “spirit of antiquity” is “already given in each individual,” i.e. can be found directly in every reading of classical works (ibid., p. 636–7), he nonetheless agrees with Ast’s principle of the interrelation between the individual and the whole, which he calls “indisputable.” On this principle, we are further told, rest “a large number of hermeneutic rules” (ibid). What these rules are supposed to regulate is, in particular, the handling of quotations and evidence, of the individual passages of a text and of their possible generalizability.

From this lecture, I now select a single segment that I understand to be representative for the ‘whole’ of Schleiermacher’s argument. This passage deals with a certain type of text passages, namely the so-called “parallel places [Parallelstellen].”

The correct procedure with parallel places is also based on the fact that one chooses only such places which occur in a whole similar to the one to be explained with respect to the word in question, and which could therefore also be parts of the same whole. But to the extent that this is not certain, the application will also be uncertain. As evident as this is and could be confirmed by several examples: as difficult is the question to answer, how far one can advance with the application of this rule. (Ibid., p. 626.)³

When Schleiermacher states that the finding of a parallelism between different places is based on the notion of similarity (“places which occur in a whole similar to the one to be explained”), the question arises whether such similarity points to an underlying identity, or whether it could also be misleading. Put differently: Do we take it as mere contingency when the same word occurs more than once in a text, or can we make sense of such repetition effects? Schleiermacher’s answer is: To be able to generate this sense hermeneutically, one has to make plausible that the parallel passages really could be “parts of the same whole.” That it could be, but it need not be. Schleiermacher therefore adds the remark “that this is not certain” and that “the application will also be uncertain.” The uncertainty in the relationship of the individual and the whole concerns the core of the hermeneutical method—and, hence, also Schleiermacher’s own methodological argumentation. He addresses this very complexity when he writes that “this,” namely, the uncertainty of application, is “evident” as a problem and “could be confirmed by several examples.” Obviously, these meta-hermeneutical examples would be negative “evidence or proofs” for the parallel passage rule: examples that could show how “difficult is the question to answer, how far one can advance with the application of this rule.”⁴

The difficulty continues in a reservation against any hermeneutic argument based on positive evidence. Schleiermacher continues that “it happens so easily that completely wrong ideas are connected with individual sentences of a writer, if the sentences are torn out of their original context only as evidence or proofs of another context” (ibid.).⁵ The concept of context as such refers to an interplay of de- and recontextualization. Therefore, sentences that can be “placed for themselves” can be placed into ever new contexts. As decontextualized they appear “indeterminate to a significant degree,” which at the same time means that they “become quite determinate only according to the context in which they are introduced” (ibid.).⁶ As much as the hermeneutic procedure aims at gaining wholeness and integral sense, it always takes into account the insistent or even irritating character of the particulars. This is to be taken account in any hermeneutic system, which is why Schleiermacher formulates his rules in such a roundabout and unspecific way. One can say that the translation of the regulatory knowledge implicit in every hermeneutic into a fixed set of rules cannot be had without the countervailing instance of the individual passage.

Accordingly, the determination of hermeneutic rules is a casuistic affair that must deal with the fundamentally “unruly” character of individual cases and examples (Gelley, 1995). For as it says immediately before the quoted passage: “in a new section, under some circumstances, other meanings can find their place with the same right as in a completely different work” (Schleiermacher, 2002, p. 626).⁷ Thus, the hermeneutic work with individual passages and their parallelization both stabilizes and destabilizes the interpreted texts. The production of context seems to guarantee the necessity of the passage chosen and to stand for the theoretical reduction of randomness. But at the same time it shows how contingent the very gesture of contextualization actually is. Every context established somewhere is a context broken open elsewhere. Every passage is a place of indeterminacy.

The ‘place of indeterminacy’ is a category of phenomenological literary theory, as conceived by the Polish philosopher Roman Ingarden, a hundred years after Schleiemacher, around 1930. According to this, literature contains “a great many places of indeterminacy [sehr viele Unbestimmtheitsstellen]” (Ingarden, 1975, p. 43).⁸ A paradigm of this is the way in which narrative texts—notably those that can be called ‘realistic’ in the broadest sense—give the impression that their characters possess a complete life into which the reader is only given some glimpses, while the larger rest remains unrepresented: “Usually whole periods of the lives of the people portrayed do not reach any explicit representation, so that the changing characteristics of these people remain indeterminate” (ibid., p. 44).⁹ In such cases, the perceived incompleteness of the text requires concretization by the reader, who “then reads, as it were, ‘between the lines,’ and involuntarily, through what may be called an ‘over-explicite’ understanding […] completes some of the sides of the representationalities presented that are not determined by the text itself” (ibid., p. 47).¹⁰ In this procedure there seem to be hardly any precautions against randomness, since the “involuntary” additions of the potentially endless number of readers of a text are, strictly speaking, methodically uncontrollable.

However, despite this localization in the reader’s mind and “between the lines,” there are repeated references in Ingarden that one can detect places of indeterminacy within the text itself. Thus, he says, there are in every literary work of art “places of the unsaid, the concealed, the indeterminate, the left open, which, despite their strange presence and their equally strange unnoticedness and unappreciatedness, nevertheless play an essential role in the artistic structure of the work of art” (ibid., p. 60–1).¹¹ Accordingly, the task of literary studies is to determine which of these places are “present,” which “may be eliminated” and which “should remain standing,” which “fillings” are possible and which are appropriate (ibid., p. 58–9). Ingarden’s concepts of eliminating and filling seem to indicate references not only to traditions of aesthetic judgment, but also to philological textual criticism, in whose procedures unclear or unreadable passages can be clarified by emendations and conjectures (cf. Bohnenkamp et al., 2010). Anyway, in such formulations he states that places of indeterminacy are not indeterminate as such, but determinable, verifiable, and locatable in the text itself. The need to be ‘filled’ is supposed to be essential to the literary phenomenology of those places. Therefore, in Ingarden, they are not the theoretical epitome of contingency, but of necessity.

It stands to reason that the West German reception aesthetics of the 1970s and 80s, particularly the school of Constance with its concept of ‘Leerstellen’ (‘empty spaces,’ ‘voids’), would refer to Ingarden. However, it did so critically and with a number of reservations. Most prominently, Wolfgang Iser noted in his book on The act of reading that Ingarden had never really shown how the finding of places and their 'concretization' was to be put into practice; instead, “he only offers banal examples when searching for a concrete illustration of how the indeterminacies are filled in” (Iser, 1978, p. 176). Iser, too, however, would have considered it banal to simply locate and indicate such places in the concrete sequence of signs in the text. Instead, he says in another paper of the same period, “Between the ‘schematized views’ a void arises from the determinacy of the views colliding with each other.” (Iser, 1975a, 1975b, p. 235)¹² For Ingarden, the term “schematized views” refers to the multiplicity of representational perspectives from which an object emerges in the literary text and each of which claims representative validity (see Ingarden, 1931, pp. 259–283).

Iser concludes from this: The more manifold the “views” of a text are, the more numerous are its voids. Hence, working with voids is not primarily a matter of a realistic mode of writing that, with its underdeterminacies, suggests that the characters have a complete (and completely determined) life somewhere outside the text. Rather, it is a matter of a genuinely post-realist overdeterminacy: “Classic examples of this would be, for instance, Joyce’s last novels, Ulysses and Finnegans Wake, where overprecision of the representational grid proportionally increases indeterminacy.” (Iser, 1975a, 1975b, p. 235)¹³ Iser thus argues for a gradation of literary complexity. The more diverse the perspectives of a text are, the more numerous are the gaps and open spaces within. Especially the ‘overprecis’ texts of literary modernism are—formulated paradoxically—full of empty spaces. Just because they no longer know contingency in their composition, they open themselves to contingency in their reception, at every point. This predetermined indeterminacy “represents the most important switching element between text and reader” (ibid., p. 248).

Iser thus seems to mark two kinds of interstices: one “between the ‘schematized views’,” another one “between text and reader.” Both, however, must in one way or another be found within the structure of the text, at least as long as literary hermeneutics will be pursued (instead of empirical reader research). The Constance school inherited Ingarden’s problem to understand empty spaces as properties of texts. All the more, the category of the void indicates a persistent problem of textual localization in reception theory (and it continues to be a problem beyond that theoretical horizon, see as recent examples Gindner, 2018; Santiáñez, 2018). Iser hints at how this could be conceived in a concrete way, linking aspects of form and of reading practice, when he speaks about the serial publications of nineteenth-century novels. Charles Dickens is a prime example of this, as he “wrote his novels only from week to week, and in between he tried to find out as much as he could about how his readers thought the plot would progress” (ibid., p. 236).¹⁴ Besides participation, serial publication leads to specific “suspense effects,” since it works with an “editing technique,” interrupting precisely “where one would like to know something about the outcome of what has just been read” (ibid., p. 237).¹⁵ From episode to episode, the reader is here “forced by the pauses prescribed to him to imagine always something more than is usually the case in continuous reading” (ibid., p. 237).¹⁶ Such cuts and gaps are then in fact concrete, specific places in the overall text of the respective novel.

This concreteness, however, remains the exception with Iser. Rather, when talking about voids, he often offers a considerable amount of negative and double-negative formulations, as in a short article in which Iser defines voids as “the omission of linguistically non-manifested connectivities,” stresses that they “do not exist as such” and speaks of their “non-givenness” (Iser, 1975a, p. 326).¹⁷ By committing the concept of the void so vehemently to negativity, Iser makes it seem almost idealistic. This is probably due to the desire not to become too ‘banal’ in terms of literary theory. In order to keep the concept of the void free from all suspicion of positivism, its concrete location is mostly avoided. Again, this is due to the theoretical interest in which Iser differs fundamentally from Ingarden: to conceive of ‘voids’ as emphatically open places in the interplay of text and reception, and thus to give contingency an essential place in the realm of aesthetic experience.

In another theoretical context of the mid-1970s, an attempt was made to positivize the category of place in an extreme way. This was the case when Jacques Derrida in his lengthy essay Limited Inc put the problem of partial or complete citation to the test. The essay stands in the context of his debate with John R. Searle on issues of citation and speech act theory: a debate that brings to view the very theory it deals with by turning discourse about speech acts itself into a series of speech acts (see Navarro Reyes, 2017). Both Searle, in his refutation of Derrida’s Signature Event Context (Signature événement contexte), and Derrida, in his reply to Searle, prominently perform the speech act of quotation. So the theoretical problems dealt with are crucially related to an “Ethics of Discussion,” as Derrida pointed out when he published his own collected contributions to the debate (Derrida, 1988, pp. 111–160). The ‘ethical’ question is how one should quote one another, or, to put it in terms of localization: in what way one grants the opponent space and places in one’s own text.

According to Derrida’s notion of iteration, quotations are conditioned by the fundamental repetitiveness of signs, a repetitiveness that at the same time contains a “force that breaks with its context” (ibid., p. 9). Further, “This force of rupture is tied to the spacing [espacement] of the written sign” (ibid.). Derrida thus understands quotation in a decidedly spatial way, as something taken out of the text (Latin: ex-emplum). Quotation is the possibility of taking something out of a text: a “possibility of disengagement” (ibid., p. 12). This possibility is always possible: a non-random possibility. Derrida insists on this over Searle’s understanding of ordinary language. But what does this mean for the mode of citation in the Derrida-Searle debate? How contingent or non-contingent are the decontextualizations and ‘disengagements’ that are performed here? In this respect, Searle is quite laconic. In his Reply to Derrida, he states at the outset that he will not address all of the numerous aspects of Signature event context: “In my reply I will not attempt to deal with all or even very many of the points he raises, but will concentrate on those that seem to me the most important and especially on those where I disagree with his conclusions” (Searle, 1977, p. 198).

Here, however, it is less a matter of dissent than of correction, as the assessments of Derrida’s reflections as misunderstandings, confusions and mistakes, repeated several times on each page, make clear. In particular, Searle finds Derrida’s interpretation of John Austin’s speech act theory so wrong that he can only name the “major misunderstandings and mistakes” in order to show the “crucial ways” of misunderstanding (ibid., p. 203). The unspoken precondition for such a statement is the conviction to be able to distinguish with certainty what is “major” and “crucial” from what is minor and irrelevant, i.e. to find the representative parts of the text. Even more: in view of Derrida’s stated confusion, it is more likely for his opponent to find the intellectual substrate beyond the text than in it. Since Searle is concerned with creating clarity and eliminating ambiguity, his reading interest is directed exclusively and explicitly to what Derrida presumably means. The doubtful, questioning gesture that determines Signature event context from the very first sentence (“Is it certain that […]?”, Derrida, 1988, p. 1) is therefore excluded from consideration as merely rhetorical; the diction, declared to be difficult to understand, must first be made usable for argumentation: “Stated in its most naked form, and leaving out the confusion about citationality, the structure of Derrida’s argument is this: […]” (Searle, 1977, p. 203). Derrida’s rebuttal is directed not least against this conviction, which he wilfully turns into its opposite. Accordingly, there is nothing at all to be left out.

On the other hand, as one will have already noticed, I do not “concentrate,” in my reading (for instance, of the Reply), either exclusively or primarily on those points that appear to be the most “important,” “central,” “crucial.” Rather, I deconcentrate, and it is the secondary, eccentric, lateral, marginal, parasitic, borderline cases which are “important” to me and are a source of many things, such as pleasure, but also insight into the general functioning of a textual system. (Derrida, 1988, p. 44)

This turn against representative places fits, it seems, precisely with Derrida’s critique of representation. However, this critique, since it articulates itself as general and fundamental, is itself dependent on a kind of representation. When Derrida, in Signature Event Context, introduces “the interpretation of writing that is peculiar and proper to philosophy,” in order to then distance himself from it, Condillac’s Essai sur l'origine des connaissances humaines suffices for him as a pars pro toto: “I shall limit myself to a single example, but I do not believe that a single counterexample can be found in the entire history of philosophy as such” (ibid., p. 3). The exemplariness of this example proves to be less in the way it is taken out of its context than in the way Derrida puts the “entire history of philosophy” into it. His treatment of Searle’s Reply is at once more subtle and more violent. It consists in an attempt to abandon representative argumentation altogether. What Derrida does to answer Searle’s objections is to quote the latter’s text not representatively but completely. Disguised as a “technical convention,” this is initially justified with the argument of reader-friendliness and clarity:

Since the readers cannot be expected to remember the two texts verbatim, and yet no resumé will be adequate, I shall quote at length, as I already said, from Sec [Signature event context] and from the Reply, in order, as far as possible, to avoid confusion, distortion, displacement, or biased selection. (Ibid., p. 45)

What Derrida does, then, in responding to Searle’s objections, is to reproduce the Reply entirely and largely in its order, but in small passages. The extent of these quotations is often not more than one word or a short phrase; only exceptionally do we find quotations of several lines. Thus, finally, Searle’s text accumulates as a whole, but it does not stand as a whole (which, in the hermeneutical tradition, would be supposed to be more than its parts). In relation to such a whole, each of these parts is rather isolated, embodying Derrida’s text-context interest with his already quoted catchwords: “eccentric, lateral, marginal.”

But the same is true for Derrida’s rebuttal. The long, hundred-page text thematizes its strict limitation outwardly and inwardly. As the title Limited Inc abc… already announces, this limitation is measured by the letters of the alphabet, which give the headings to the sections of the essay. The first section of the text is headed “d” (because “abc” is already in the title). Up to and including section “k,” Derrida deals with theoretical preliminaries and “technical conventions.” The actual discussion with Searle thus takes place between “l” and “z.” The letters refer to the order of the alphabet, which is known to be arbitrary. Derrida, however, relates arbitrariness and contingency: “This is not without a certain arbitrariness. Its effect: henceforth I will have at my disposal only 18 letters or 18 blows and I will have to make the best of them. But, one will protest, is not this limit utterly contingent, artificial and external? […] That is the question” (ibid., p. 45).

One must add, however, that while the characters of the alphabet are arbitrary, the positions of the individual letters in the alphabet are conventionally clearly fixed and therefore precisely non-contingent. In this way, Derrida condemns his own text as well as Searle’s to radical immanence and performs, indeed executes, the process of decontextualization and iteration in a manner that is not exemplary but totalizing. The opponent is completely present with his text, and because there is no remainder of his text, he can be completely brought down. Therefore, Derrida’s response to Searle’s Reply is ultimately a reading that aurates, establishes presence, and radically excludes contingency.

To conclude: The way in which literary theory handles a given text—the location of relevant places, their declaration as somehow representative for the text as a whole, and their re-integration into a new context of theoretical argument—is always afflicted with contingency. This applies to the actual selection of the passages, to the decision on their exact cut (how is a ‘place’ to be singled out; where does a ‘passage’ begin and end) and to the way in which their exemplariness is made plausible. None of these parameters is ever necessary in a strictly logical sense; one could always make other decisions. This difficulty can be understood as ‘only’ methodical, namely as an unavoidable element of randomness in the procedures of literary theory. But it can also be seen as a theoretical problem in its own right: a problem of theorizing chance.

I have outlined three different ways of dealing with this matter: hermeneutic, reception-aesthetic, and deconstructivist. In Schleiermacher’s hermeneutics, the individual text passages are at the same time decisive reference points and disturbing variables. It is this contradiction that discloses the fundamental ambivalence of the hermeneutical process, the indecision between focusing on the ‘whole’ or the ‘parts.’ Compared to this, Ingarden’s idea of the empty spaces to be filled in reception seems less susceptible to disruption, but also somewhat reductionist. In Iser’s reading, this tendency to abbreviate the complexity is criticized and the contingencies of textual localization are again emphasized more strongly. Finally, in the Searle-Derrida debate, the arbitrariness of locating relevant passages becomes the very subject of theoretical discussion, while the space of the citable text turns into a theoretical combat zone. Thus it turns out that localization and its contingencies are one of the often overlooked but particularly controversial problems of literary theory.